Chapter 8: The Investigation of Real Fires project

Introduction

¶8.1 Between 1988 and 2017 the government engaged BRE to monitor the effectiveness of the Building Regulations and the guidance contained in Approved Document B by examining the circumstances and causes of real fires. The work was known as the "Investigation of Real Fires" project.⁸²⁸⁸²⁸ {BRE00001054/6}; Shipp {BRE00047594/24} page 24, paragraph 118. It was carried out under a series of contracts between BRE and DCLG (the department),⁸²⁹⁸²⁹ Martin {CLG00019469/27} page 27, paragraph 86; {BRE00001054/6}. each of which ran for about two or three years and was based on an invitation to tender, and a proposal submitted by BRE for the work.⁸³⁰⁸³⁰ Crowder {Day229/126:11-14}; {HOM00046292}; {CLG10003892}.

Background

¶8.2 BRE, originally known as the Fire Research Station, had been involved in the investigation of fires since the early 1970s. In 1970 the Fire Survey Group was formed, which consisted of a number of scientists who attended the scene of selected fires.⁸³¹⁸³¹ Shipp {BRE00047594/21} page 21, paragraph 108. In 1988 Martin Shipp took over as head of the Fire and Security Section of the Fire Research Station, which by then was the department with responsibility for carrying out investigations into fires.⁸³²⁸³² Shipp {BRE00047594/3} page 3, paragraph 14; Shipp {BRE00047594/21} page 21, paragraph 108; Shipp {BRE00047594/24} page 24, paragraph 118. In that capacity he attended the scenes of fires, drafted reports for the department and engaged with departmental officials.⁸³³⁸³³ Shipp {BRE00047594/24} page 24, paragraph 118. Thereafter, he was involved in the investigation of a number of significant incidents, such as the fire at the Yarl's Wood Detention Centre in 2002 and the fire at the Rose Park Care Home in 2004.⁸³⁴⁸³⁴ Shipp {BRE00047594/5} page 5, paragraphs 27 and 28; {CLG10003892/1} paragraph 1.

¶8.3 The purpose of the Investigation of Real Fires project was to ensure that the government was informed in good time of new matters affecting fire safety, particularly those that emerged from significant incidents, and thereby to inform government policy.⁸³⁵⁸³⁵ Shipp {BRE00047594/26} page 26, paragraph 127; {CLG10003892/3}. More particularly, the principal objective was to identify whether there were aspects of the guidance in Approved Document B or the functional requirements in Part B of Schedule 1 to the Building Regulations that were not working as intended and required amendment.⁸³⁶⁸³⁶ Crowder {Day229/94:23}-{Day229/95:11}; Crowder {BRE00047668/12} page 12, paragraph 51; Shipp {BRE00047594/26} page 26, paragraph 127; Martin {CLG00019469/27} page 27, paragraph 86. The project had been established following the coming into force of the Building Act 1984 in recognition of the fact that it was necessary to monitor the introduction of the functional requirements to ensure that they were operating as envisaged.⁸³⁷⁸³⁷ Crowder {Day229/95:13-24}. Dr David Crowder was involved in the project at BRE from 2006 onwards and took over responsibility for running it from April 2015.⁸³⁸⁸³⁸ Crowder {BRE00047668/6} page 6, paragraph 27; Crowder {Day229/96:1-7}.

122

¶8.4 The project was not concerned with matters that had no potential bearing on the statutory guidance or the Building Regulations, such as fires started deliberately and accidental fires resulting from a simple failure to follow the regulations or guidance.⁸³⁹⁸³⁹ Crowder {BRE00047668/12} page 12, paragraph 51; Crowder {Day229/101:6-23}. By 2006 the following objectives for the project had been identified:⁸⁴⁰⁸⁴⁰ {HOM00046642/13} section 2. a. To provide timely reports to the department on the effectiveness of the guidance in

¶Approved Document B in achieving fire safety in buildings in England and Wales. b. To improve understanding of how unusual fires develop and grow, particularly in domestic and other residential properties. c. To monitor the effect of European standards on building materials and systems. d. To monitor the approach adopted by fire safety engineering and other means of meeting the requirements of the Building Regulations. e. To identify the need for research relating to specific problems identified in fire investigations. f. To identify the efficacy of the Building Regulations in providing protection for fire fighters. g. To maintain close contacts with investigators, including the fire and rescue services, to

encourage the exchange of information on unusual fires that would be of benefit to the department.

¶h. To disseminate findings from fire investigations to the fire and rescue services, designers and owners of buildings and others, as appropriate.

¶8.5 From 2007 onwards Dr Debbie Smith was BRE's contract manager for the project and remained responsible for it at the time of the Grenfell Tower fire.⁸⁴¹⁸⁴¹ Shipp {BRE00047594/23} page 23, paragraph 117; Smith {MET00081237/45} page 45, paragraph 179; Smith {Day236/159:22-25}; Day236/160:24}-{Day236/161:5}. However, she did not have day to day involvement in it, although she was involved in any formal correspondence with the department.⁸⁴²⁸⁴² Crowder {Day229/98:18}-{Day229/99:20}; Smith {MET00081237/45} page 45, paragraph 179; Lennon {BRE00043688/9} page 9, paragraph 48. Most of the reports produced pursuant to the project after 2007 were approved by Dr Corinne Williams, deputy contracts manager for the programme, acting on her behalf. Each report underwent technical checks or peer review before being sent to Dr Williams or Dr Smith for approval.⁸⁴³⁸⁴³ Holland {BRE00043829/10} page 10, paragraph 33; Crowder {Day229/99:22}-{Day229/100:18}.

¶8.6 The department regarded the project as a means of providing robust evidence to ensure that the Building Regulations and Approved Documents were effective and continued to be fit for their purpose.⁸⁴⁴⁸⁴⁴ Martin {CLG00019469/27-28} pages 27-28, paragraph 86. It was seen as a means of enabling the department to take an active approach to ensuring that policies and guidance were proportionate to the risk and gain a greater understanding of industry practice.⁸⁴⁵⁸⁴⁵ Martin {CLG00019469/28} page 28, paragraph 87(c) and (d). The information obtained from the project, together with the statistics it was receiving, led to a general understanding within the department that Approved Document B provided adequate guidance on the use of combustible materials.⁸⁴⁶⁸⁴⁶ Burd {CLG00019461/25} page 25, paragraph 64.

123

The operation of the project

¶8.7 In practice, the project involved monitoring news reports of fires, from which BRE selected incidents to include in its reports. Important sources of information were media reports or local fire and rescue service reports,⁸⁴⁷⁸⁴⁷ Shipp {BRE00047594/25} page 25, paragraph 123; Crowder {Day229/122:10}-{Day229/123:24}. both of which were monitored on line.⁸⁴⁸⁸⁴⁸ Crowder {Day229/121:7-23}. A list of high – and low-priority considerations were agreed with the department at the beginning of each contract and was used by BRE to determine whether a more detailed examination of any particular incident was required.⁸⁴⁹⁸⁴⁹ Shipp {BRE00047594/25} page 25, paragraph 124. At the top of the high-priority list were fatal fires, but the list also included fires of relevance to the Building Regulations or Approved Document B and fires of particular scientific interest, including fires involving cladding.⁸⁵⁰⁸⁵⁰ Crowder {BRE00047668/41-43} pages 41-43, paragraph 171; Crowder {Day229/129:12}-{Day229/131:16}. Included in the low-priority list were fires overseas, a subject to which we return below.⁸⁵¹⁸⁵¹ Crowder {BRE00047668/41-43} pages 41-43, paragraph 171. The list of priorities did not fundamentally change during the life of the project⁸⁵²⁸⁵² Crowder {Day229/131:17-22}. and largely determined the way in which BRE responded to or investigated any particular incident.⁸⁵³⁸⁵³ Crowder {Day229/124:18}-{Day229/126:3}; Williams {BRE00043695/10} page 10, paragraph 43; Shipp {BRE00047594/25} page 25, paragraphs 124-125.

¶8.8 If a site inspection was considered necessary, BRE contacted the relevant fire and rescue service in order to gain access to the scene, assuming that the department was content for it to do so.⁸⁵⁴⁸⁵⁴ Shipp {BRE00047594/25} page 25, paragraph 125. Sometimes BRE was unable to obtain access to the scene of a fire, for example, if the premises were a crime scene.⁸⁵⁵⁸⁵⁵ Shipp {BRE00047594/25} page 25, paragraph 126.

¶8.9 BRE produced a range of different reports for the department based on its investigations, including quarterly progress reports, year-end reports, research reports at the end of each two or three-year contract cycle and reports on fires of special interest if site investigations had been carried out.⁸⁵⁶⁸⁵⁶ Holland {BRE00043829/17-18} pages 17-18, paragraph 51 (category 1 fires being identified at Holland {BRE00043829/11} page 11, paragraph 37); Crowder {Day229/102:3}-{Day229/103:2}. BRE also wrote articles for various publications based on its work.⁸⁵⁷⁸⁵⁷ Holland {BRE00043829/17-18} pages 17-18, paragraph 51. Some experimental fire testing was carried out under the project.⁸⁵⁸⁸⁵⁸ Martin {CLG00019469/27-28} pages 27-28, paragraph 86.

¶8.10 BRE typically produced a few hundred reports on fires each year, although the number varied in accordance with the quality of the information available and the type of fires that had occurred. For example, in the year from July 2010 to July 2011 204 reports were produced, whereas 511 were produced in the year from July 2011 to July 2012. The budget for the project allowed for 28 days a year to be spent on preparing reports, which included monitoring news reports and sifting out fires that were not worth reporting. In practice three quarters of the time was spent on monitoring and one quarter on writing reports. That amounted to about 15 minutes a fire.⁸⁵⁹⁸⁵⁹ Crowder {BRE00047668/43} page 43, paragraph 173. The contract for the years 2012 to 2015 allowed for a maximum of eight site visits a year.⁸⁶⁰⁸⁶⁰ Holland {BRE00043829/14-15} pages 14-15, paragraph 42; {BRE00000951/21} under heading "Category 1 incidents". The total budget for the project remained at around £100,000 per year between 2006 and 2017.⁸⁶¹⁸⁶¹ Crowder {BRE00047668/44-45} pages 44-45, paragraph 179; Crowder {Day229/145:16}-{Day229/146:3}.

124

¶8.11 At the start of each contract a letter was sent to all fire and rescue services asking for their co-operation with the project and describing the list of high and low priorities that had been agreed with the department so that they could understand which fires BRE was interested in.⁸⁶²⁸⁶² Crowder {Day229/126:25}-{Day229/127:13}; {BRE00011186}.

A change in the scope of the project: October 2012

¶8.12 The department changed the terms of the contract for the period November 2012 to March 2015, which was put out to tender on 3 September 2012.⁸⁶³⁸⁶³ {BRE00027677/1} paragraph 1(ii). Those changes were reflected in BRE's proposal dated 1 October 2012.⁸⁶⁴⁸⁶⁴ {BRE00000951/4}. That proposal, which was prepared by Martin Shipp and approved by Dr Smith,⁸⁶⁵⁸⁶⁵ {BRE00000951/2}. was submitted to Brian Martin at the department.⁸⁶⁶⁸⁶⁶ {BRE00000951/1}. The overarching objectives of the project included the provision of unbiased, robust and independent evidence and information that would enable the department to ensure that the guidance in Approved Document B remained fit for purpose and to support the review of policy more generally.⁸⁶⁷⁸⁶⁷ {BRE00000951/4}. They also included identifying areas for research and for potential changes to Part B of Schedule 1 of the Building Regulations or Approved Document B arising from specific problems identified in investigations.⁸⁶⁸⁸⁶⁸ {BRE00000951/4}.

¶8.13 The proposal also contained a section describing the way in which reports would be produced. It included the following provision, which reflected a requirement that had been introduced for the first time in the department's request for a proposal:⁸⁶⁹⁸⁶⁹ {BRE00000951/51}; Crowder {BRE00047668/12-13} pages 12-13, paragraph 53.

"All reports will: … • Not contain any policy recommendations. All reports produced by BRE for this project will not contain any proposed text for a revision to an Approved Document or supporting guidance. With the agreement of, or at the request of, DCLG, such material will only be provided in a separate Policy Implications Report (as and when necessary…) as this will typically need to be published separately."⁸⁷⁰⁸⁷⁰ {BRE00000951/51} under heading "All reports will:".

¶8.14 The contract for the project was awarded to BRE on 8 November 2012.⁸⁷¹⁸⁷¹ {BRE00027677}. At a meeting held at the start of the contract between Martin Shipp, Ciara Holland and Dr Crowder for BRE and Brian Martin and Steve Kelly for the department, the department made it clear that BRE should not volunteer policy recommendations but provide them only if asked to do so.⁸⁷²⁸⁷² Crowder {Day229/110:10-19}. Dr Crowder regarded that as a fundamental change in the way in which the contract operated⁸⁷³⁸⁷³ Crowder {Day229/113:14-15}. and thought it was a result of the government's deregulatory policies.⁸⁷⁴⁸⁷⁴ Crowder {Day229/114:1-3}; {Day229/118:4-10}. That shift in approach was entirely consistent with the evidence we heard from senior civil servants in the Department who explained to us the pressures they were under at the time due to the government's overarching desire to deregulate.⁸⁷⁵⁸⁷⁵ As discussed in more detail in Chapter 10. As far as we know, the department did not ask BRE to produce any reports on policy implications.⁸⁷⁶⁸⁷⁶ Crowder {Day229/119:3-13}; Smith {Day236/170:12-25}.

125

Findings in reports

¶8.15 Each of the final research reports produced by BRE for the department between 2001 and 2015 pursuant to the project, contained the following conclusion:

"The findings from this period have reaffirmed the overall effectiveness of the building regulations and ADB in providing for the safety of life in the event of fire and most of the significant issues that have been identified during this study fall outside the scope of these regulations."⁸⁷⁷⁸⁷⁷ {BRE00000936/15} section 3.

¶8.16 Although Martin Shipp said that the paragraph correctly reflected the position, there were periods during which it self-evidently did not.⁸⁷⁸⁸⁷⁸ Martin {Day251/174:1-8}; {Day251/174:20}-{Day251/175:1}. One striking example is the report dated 31 March 2005,⁸⁷⁹⁸⁷⁹ {CLG00019455/1}. which included a summary of the fire at The Edge, Salford, on 6 January 2005.⁸⁸⁰⁸⁸⁰ {CLG00019455/76-79} paragraph 4.1.148. It contained the same standard paragraph.⁸⁸¹⁸⁸¹ {CLG00019455/3}. However, the fire at The Edge had not only raised obvious questions within the department at the time, but had shown that the guidance in Approved Document B was not effective in restricting the use of external wall panels with combustible cores on high-rise buildings,⁸⁸²⁸⁸² Martin {Day251/175:2-9}; {Day251/176:1-5}. which had led in turn to a significant amendment. In those circumstances the inclusion of that paragraph is inexplicable.⁸⁸³⁸⁸³ Martin {Day251/175:10-16}; {Day251/175:21-25}.

Sudbury House and Taplow House fires: 2010–2012

¶8.17 Two fires that featured in the reports prepared by BRE were those that occurred at Sudbury House in Wandsworth in August 2010 and Taplow House in Swiss Cottage in January 2012.

¶8.18 A report on the fire at Sudbury House was included in the report for the period July 2010 to July 2011 dated 27 July 2011.⁸⁸⁴⁸⁸⁴ {BRE00000945}. It contained the following account:

"Block of flats, London, 1st August 2010 Sudbury House, Wandsworth High Street, London. A fire occurred on the fifth floor of a 24-storey block of flats at around 11.00. A man, presumed by BRE Fire and Security fire investigators to have been in the flat of origin, was suffering from smoke inhalation and was rescued from the fifth floor and another person was rescued from the third floor. The cause of the fire was under investigation."⁸⁸⁵⁸⁸⁵ {BRE00000945/33} paragraph 5.1.3.

¶BRE did not carry out an investigation at the site of the fire;⁸⁸⁶⁸⁸⁶ Crowder {Day229/160:15-17}. the summary was prepared solely on the basis of news reports.⁸⁸⁷⁸⁸⁷ Crowder {Day229/160:18-20}. However photographs taken later showed that a fire in the compartment of origin had caused the building's cladding to catch fire. The fire had spread vertically through the cladding and up a column and had clearly gone beyond the compartment of origin.⁸⁸⁸⁸⁸⁸ Crowder {Day229/161:18-20}. The rainscreen at Sudbury House was replaced after the Grenfell Tower fire because it also consisted of Aluminium Composite Material (ACM) panels with a polyethylene core.⁸⁸⁹⁸⁸⁹ Crowder {Day229/161:23}-{Day229/162:3}.

126

¶8.19 The fire at Taplow House was included in the report for the period July 2011 to July 2012 dated 26 July 2012.⁸⁹⁰⁸⁹⁰ {BRE00000947/53} paragraph 5.1.64. It contained the following account:

"Block of flats, London, 16th January 2012 Swiss Cottage, London. The fire started at about 22:00 on the 17th floor of a 22-storey block of flats. 130 people were evacuated to a nearby community centre. The 17th floor was partially gutted, but fire fighters confined the fire to that floor and brought the fire under control in over three hours. The fire was caused by a candle setting fire to papers, prompting a warning by a London Fire Brigade spokesman on storing large amounts of papers, magazines and books in dwellings."

¶Again, BRE did not visit the scene of the fire to carry out an investigation; the summary was based on media reports.⁸⁹¹⁸⁹¹ Crowder {Day229/163:8-15}. However, the fire spread both upwards and downwards over one column of the tower.⁸⁹²⁸⁹² Crowder {Day229/164:4-6}. After the Grenfell Tower fire the cladding system at Taplow House was found to consist of Reynobond 55 ACM PE rainscreen panels with mineral wool insulation.⁸⁹³⁸⁹³ Crowder {Day229/163:16-19}.

¶8.20 Although the fire at Sudbury House had required the rescue of two people, which suggested that it was a high-priority for investigation,⁸⁹⁴⁸⁹⁴ The second category under the high-priority list was "Non-fatal fires with injury or rescue": Crowder {BRE00047668/41-43} pages 41-43, paragraph 171. there was nothing in BRE's summaries of either of those fires that suggested that they ought to be the subject of further investigation.⁸⁹⁵⁸⁹⁵ Crowder {Day229/168:18-23}; {Day229/169:13-22}. It seems that BRE did not regard fires in high-rise buildings as in themselves requiring further investigation, because they were not expected to spread beyond the compartment of origin.⁸⁹⁶⁸⁹⁶ Crowder {Day229/170:2-22}. The fire at Sudbury House was not regarded as a cladding fire comparable to Knowsley Heights, Garnock Court or Lakanal House, since it had been largely contained in the dwelling of origin.⁸⁹⁷⁸⁹⁷ Crowder {Day229/170:22}-{Day229/171:1}.

¶8.21 The approach to these two fires exposes a fundamental flaw in the way the Investigation of Real Fires project was conducted. BRE's research was based on very limited information (mainly media reports), with the result that it had no reliable way of telling whether a particular fire required more detailed investigation. It may be that the quality of the information available to BRE was declining over time, either because of reduced funding by the department or because it was increasingly dependent on the media rather than information from fire and rescue services.⁸⁹⁸⁸⁹⁸ Crowder {Day229/171:3-20}. In our view, however, it was unsatisfactory for the department and BRE to be pursuing a project of this kind on the basis of such superficial information. BRE ought to have told the department that its findings and advice were becoming increasingly unreliable and that unless the budget could be improved to maintain standards it could no longer provide information of the quality the department expected to receive. For its part, the department should have considered whether the project was any longer meeting its objectives, and if not, why not.

127

¶8.22 In fact, there was additional information available in the public domain about the fire at Sudbury House that would have made it clear to BRE that the incident involved a cladding fire that deserved further investigation.⁸⁹⁹⁸⁹⁹ Bisby, Phase 2 Report {LBYP20000001/213} paragraph 1210. In a YouTube video posted online on 1 August 2010 (the same day as the fire) flames could be seen spreading up the column (and to a lesser extent down it) and it was clear that the fire had been extinguished only when firefighters had been able to bring water directly to bear on it.⁹⁰⁰⁹⁰⁰ See stills from the video {INQ00015117/3-4}. The title of the video was "Fire on the 5th floor, Sudbury House, Wandsworth High Street" and includes a link to a BBC article about the fire.

¶8.23 BRE ought to have found that video.⁹⁰¹⁹⁰¹ Crowder {Day229/178:15-25}. The incident looked like a cladding fire of the kind that the department would wish to have been aware of⁹⁰²⁹⁰² Crowder {Day229/179:4-9}. and more should have been done by BRE by way of investigation.⁹⁰³⁹⁰³ Crowder {Day229/179:11-21}. In our view BRE did not pursue the project with the vigour that was required, but the project itself was flawed in a number of respects,⁹⁰⁴⁹⁰⁴ Crowder {Day229/180:25}-{Day229/181:2}. not least because the funding provided for the work did not increase between 2006 and 2017 to match the increased cost of carrying out the work.⁹⁰⁵⁹⁰⁵ Crowder {Day229/180:3-24}.

Identification of patterns and trends

¶8.24 Another weakness affecting the project was the existence of a belief on the part of the department and BRE that the findings from any particular investigation could not be of general significance⁹⁰⁶⁹⁰⁶ Shipp {BRE00047594/27-28} pages 27-28, paragraph 134. and that it would therefore be necessary to identify patterns or trends before drawing the department's attention to developments that might call for action. Each annual report began with a table⁹⁰⁷⁹⁰⁷ {HOM00046642/16-29}; {BRE00000945/19-30}; {BRE00000947/25-43}. that identified aspects of interest from which BRE attempted to identify trends that required specific identification in the report.⁹⁰⁸⁹⁰⁸ Crowder {Day229/147:8-13}. Individual incidents, however serious, did not influence the conclusions in the end of year reports because no pattern had been identified.⁹⁰⁹⁹⁰⁹ Crowder {Day229/147:25}-{Day229/149:9}. BRE was interested in whether a problem had become endemic;⁹¹⁰⁹¹⁰ Crowder {Day229/149:25}-{Day229/150:20}. a single incident would need to be combined with information from other events to identify trends that satisfied the objectives of the project.⁹¹¹⁹¹¹ Crowder {BRE00047668/44} page 44, paragraph 177.

¶8.25 That was a very naïve approach, both on the part of BRE and the department. Although the identification of patterns or common occurrences was no doubt an important part of BRE's work, it failed to accord sufficient recognition to the fact that a single rare but significant incident might well have important implications for the effectiveness of the regulatory regime. It might take only one serious fire to demonstrate that the guidance was inadequate. The fire at Grenfell Tower was but one example in a long line that included Knowsley Heights, Garnock Court and Lakanal House. The fundamental shortcoming of the Investigation of Real Fires project was that it concentrated on patterns based on the frequency rather than the severity of incidents. In that way the project undermined rather than enhanced the ability of the government to make a proper assessment of fire risk in the built environment. BRE ought to have made that clear to the department.

128

Concerns about downward fire spread and flaming droplets

¶8.26 The narrowness of the approach taken to the identification of patterns or trends was particularly apparent in the way in which project reports addressed the subject of downward fire spread. Dr Crowder was concerned about the absence of any express provision in Approved Document B relating to downward fire spread.⁹¹²⁹¹² Crowder {BRE00043716/25} page 25, paragraph 82. Between December 2007 and August 2016 there had been a number of incidents in which there had been downward spread of fire as a result of the building fabric, all of which had been individually reported to the department.⁹¹³⁹¹³ Crowder {BRE00043716/23-24} pages 23-24, paragraph 79 (a) to (g). Materials that formed burning droplets were often fitted against other flammable materials, such as combustible insulation. Burning droplets would typically promote the rapid spread of fire by igniting other combustible materials more quickly than would otherwise have been the case.⁹¹⁴⁹¹⁴ Crowder {BRE00043716/25} page 25, paragraph 83.

¶8.27 Despite Dr Crowder's concerns, none of BRE's reports identified the creation of burning droplets or the downward spread of fire as matters that deserved to be brought to the department's attention as potential reasons for reviewing the statutory guidance. That appears to have been because the incidents had occurred sporadically over a 10-year period⁹¹⁵⁹¹⁵ Crowder {Day229/214:13-23}. and a trend had not been identified in any one report.⁹¹⁶⁹¹⁶ Crowder {Day229/217:7-11}. Dr Crowder thought that such incidents could have been attributable to other causes.⁹¹⁷⁹¹⁷ Crowder {BRE00043716/25} page 25, paragraph 82. There was, therefore, little or no effort to identify trends that might extend across separate reporting periods. BRE made it clear to the government that it needed to look at innovative construction products and techniques from a fire safety perspective⁹¹⁸⁹¹⁸ Crowder {Day229/214:23}-{Day229:216:2}; {Day229:217:12-18}. but the government gave the impression that it was unwilling to make changes.⁹¹⁹⁹¹⁹ Crowder {Day229/208:23}-{Day229/209:9}.

¶8.28 We accept that the government's enthusiasm for deregulation after 2010 made it reluctant to receive advice that the regulatory regime was inadequate, but it is regrettable that BRE did not give clear advice about what could be learnt from individual incidents or about trends that were evident over a number of years, such as with the danger posed by burning droplets and their contribution to the downward spread of fire. At no stage did BRE offer to report to the department on the policy implications of the downward spread of fire and the adequacy of Approved Document B to address it.⁹²⁰⁹²⁰ Crowder {Day229/209:1-9}. Instead, BRE appears to have decided not to give advice to the government that it knew would be unwelcome.

International fires

¶8.29 Another area of importance that received limited attention from BRE was international fires, including international cladding fires. Overseas fires were identified by the department as a low priority for the project.⁹²¹⁹²¹ Crowder {BRE00047668/41-43} pages 41-43, paragraph 171. However, it would have been useful for BRE to obtain better information about international fires, such as the various fires involving ACM panels that occurred around the world and were prominently reported in the five or so years before the fire at Grenfell Tower.⁹²²⁹²² Crowder {BRE00047668/47} page 47, paragraph 184. Dr Crowder had been able to attend the scene of only one overseas fire as part of his work on the project, a fire in Dublin in 2008 that was relevant to the spread of fire in car parks.⁹²³⁹²³ Crowder {BRE00047668/47} page 47, paragraph 184. Although some international fires were mentioned in BRE's reports, very little detail was given about their causes. For example, two cladding fires in Dubai that occurred in 2016 were referred to in the end of year report for the period April 2016 to March 2017.⁹²⁴⁹²⁴ {BRE00000959/111-113} paragraphs 6.14.6 and 6.14.17 – the two fires were the Sulafa Tower fire in Dubai on 20 July 2016 and the Palm Jumeriah fire in Dubai on 12 December 2016. Although a brief description of each fire was provided, no information was given about what materials were thought to have been involved, even though reports available in the UK indicated that they had included ACM panels.⁹²⁵⁹²⁵ https://www.dailymail.co.uk/news/article-3699272/Fire-breaks-luxury-75-storey-tower-Dubai.html (20 July 2016). Nor was there any discussion in the report about the spread of fire downward, which was a feature of both fires. Mr Martin said that he had asked BRE to use its contacts to obtain more information about some of the international cladding fires that were being reported, but that the Investigation of Real Fires project was limited by the resources allocated to it.⁹²⁶⁹²⁶ Martin {Day252/118:2-6}.

129

¶8.30 At no stage during the time that Dr Crowder was involved in the project was any proposal, whether formal or informal, put to the department by BRE recommending that closer attention should be given to international cladding fires.⁹²⁷⁹²⁷ Crowder {Day229/141:22-25}. The prevailing view was that the incidents had occurred as a result of inadequate regulation and that there was no cause for concern in the UK,⁹²⁸⁹²⁸ Crowder {Day229/141:4-21}. which had higher regulatory standards. It was also thought within the department that ACM panels were not being installed in the UK.⁹²⁹⁹²⁹ Crowder {Day229/135:2-14}; Smith {MET00081237/21} page 21, paragraph 83; Martin {Day255/108:7-12}. That was obviously wrong. The department could have discovered by July 2014 that ACM cladding panels with a polyethylene core were in widespread use and that many people in the industry thought that if a panel had a Class 0 surface it could be used at any height. After early 2016 Mr Martin could have been in no doubt about that.

¶8.31 Although we recognise the significance of hindsight, we find it surprising that more was not done to monitor international fires, particularly international cladding fires. A spate of such fires occurred in the UAE in 2012 and 2013 and a further spate in 2015 and 2016. It is a matter of concern that they do not appear to have prompted any serious consideration within BRE or the department of whether closer attention should be paid to them or to the lessons they might offer. Had they done so, there might have been a more thorough investigation of what had caused such a catastrophic spread of fire across the walls of the building in each case. That might in turn have shaken the department out of its complacent belief that it could not happen here. Both the department and BRE were warned on a number of occasions about the problems that could be caused by the use of combustible materials in the external walls of tall buildings with specific reference to cladding fires that had occurred abroad. Those warnings appear to have generated at best some informal conversations between BRE and the department, but more could and should have been done.

130

Conclusions

¶8.32 In a number of important respects the Investigation of Real Fires project was flawed. Most of the reporting amounted to little more than reciting information about fires obtained from news reports, followed by formulaic conclusions that assured the department that the regulations and guidance were effective, without their efficacy having been subjected to any proper scrutiny. There was no analysis of the lessons to be learnt from significant single incidents, from the identification of patterns across different reporting periods, or from significant fires overseas. The work was being carried out at such a high level of generality that it would have been difficult for BRE to identify any patterns indicating that changes to the statutory guidance were necessary to ensure that it remained relevant to the risks posed by the built environment. From 2012 the BRE was hobbled in its reporting by a prohibition on making policy recommendations involving changes to the regulations or guidance. It knew that the government had no appetite for further regulation and therefore decided not to offer unwelcome advice. In our assessment the operation of the project epitomised what had gone wrong in BRE's relationship with the government; what was needed was proper independent advice, which the department did not want to receive. The project thus helped to foster an attitude of complacency within both BRE and the government about the adequacy of Approved Document B. We are critical of both BRE and the department for allowing that state of affairs to persist for so many years.

131